Field evaluation of drone and AI assisted larval source management in Ghana.

Abstract

Malaria remains a major public health burden in sub-Saharan Africa. In Ghana, in particular, larval source management (LSM) is increasingly recognized as a complementary vector control strategy. This study evaluates a field-adapted LSM approach that integrates drone-based mapping and artificial intelligence (AI)-driven site prioritization to enhance operational efficiency and reduce resource use. The intervention replaces conventional manual scouting with aerial mapping conducted one day prior to larvicide application. An AI model analyzes geospatial and morphological features of water bodies to identify high-risk larval habitats. Site coordinates are transmitted to field teams via mobile devices for targeted treatment. A comparative field trial was conducted in eight administrative sub-districts within Ghana's Eastern Region. Four sub-districts implemented the drone- and AI-assisted approach, while four served as controls using standard LSM procedures. A mixed-methods evaluation was employed, incorporating quantitative metrics and qualitative field insights. Drone-assisted mapping led to more than a threefold increase in the number of identified breeding sites. AI-based targeting reduced larvicide consumption by over 60%. The combined technologies lowered worker requirements by approximately 50%. Despite these reductions, malaria case trends in the intervention sub-districts remained comparable to those in the control sub-districts. The study's limitations include its restriction to the dry season and below-average rainfall, which may have influenced mosquito abundance and transmission. Drone- and AI-assisted LSM demonstrated substantial resource savings without compromising vector control outcomes. Further longitudinal evaluation across transmission seasons is warranted to assess sustained effectiveness and inform national policy.